1. FIELD OF THE INVENTION
The present invention relates to a digital signal reproduction apparatus for reproducing digital signals recorded on a recording medium such as a magneto-optical disc.
2. DESCRIPTION OF PRIOR ART
A magneto-optical disc has a magnetic thin film as a recording medium on the surface of the disk which has a magnetic anisotropy such that the axis of easy magnetization is oriented vertical to the surface of the film. Recording data onto the magnetic thin film or reproducing data from the magnetic thin film is performed by irradiating a laser beam, condensed into a diameter of about 1 .mu.m, onto the film.
When recording data signals, a strong laser beam is irradiated onto the thin film. Due to this irradiation, the temperature of the irradiated spot on the thin film is raised locally to decrease the coercive force. Therefore, when a biased magnetic field is applied to the spot, the magnetization of the spot is inverted. Accordingly, if the strength of the laser beam or the direction of the biased magnetic field is controlled according to the data signals to be recorded, they are vertically recorded onto the thin film as dots having a diameter of the laser beam. When erasing the recorded signals, the method is substantially the same as the recording method.
Upon reproducing the data signals, a weaker laser beam than the recording beam is irradiated onto the magnetic thin film. When the laser beam irradiates a spot on the film, it is linearly polarized to have a polarization plane inclined according to the magnetized state of the spot by the magneto-optical effect (Faraday effect or Karr effect) and is reflected from the surface of the film. Therefore, if the inclination of the polarization plane can be detected by converting it into an electrical signal by an optical detector through an analyzer, the recorded data signal can be reproduced.
FIG. 3 is a schematic diagram of the reproduction system for the magneto-optical disc.
When a laser beam generated by an optical head 1 is irradiated onto the thin film of the magneto-optical disc 2, the irradiated beam is reflected toward the optical head 1. The reflected light beam is converted into a reproduction signal corresponding to the inclination of the polarization plane. The reproduction signal is transmitted to a data detecting circuit 3 and the recorded data signal is outputted as a reproduction data signal.
The conventional data detecting circuit 3 used for the above reproduction system will be described referring to FIGS. 4 and 5. As shown in the upper most row of FIG. 5, if the recording data signal Sll having a series of "0"and "1" is recorded on the thin film of the magneto-optical disc, the reproduction signal S12 generated from the recording data signal S11 has a waveform wherein each peak corresponds to "1" of the recording data signal Sll. The reproduction signal S12 is inputted into a differential circuit 21, a comparative voltage generating circuit 22, a comparator 23 of the data detection circuit 3, respectively, as shown in FIG. 4. The differential circuit 21 differentiates the reproduction signal S12 and outputs a differential signal S13. The differential signal S13 is a waveform having zero-cross points corresponding to individual maximum and minimum values of the reproduction signal S12 as shown in FIG. 5. The differential signal S13 is inputted into a comparator 24. The comparator 24 compares the differential signal with a 0 V level and outputs a differential digital signal S15. Each trailing position of pulses of the differential digital signal S15 corresponds to a zero-cross point of the differential signal S13 at which it changes from a positive value to a negative value, and therefore, it corresponds to each peak point (maximum point in, the conventional example) of the reproduction signal S12. Namely, it represents a position corresponding to "1" of the recorded recording data signal S11. Therefore, if the trailing position is detected by a trailing detecting circuit 25, a peak detecting signal corresponding to "1" of the recording data signal S11 can be obtained.
However, when "0"continues in the recording data signal Sll, and therefore, the reproduction signal S12 is held at the minimum level for a while, the differential signal S13 is held at the zero-cross level for a relatively long time as indicated by a circle E of FIG. 5. In this state, noises are apt to cause in the differential digital signal S15 and the peak detecting signal S17. Therefore, it becomes necessary to remove the noise contained in the peak detecting signal S17 in order to pick up only pulse representing the peak point of the reproduction signal S12. In order for this to be realized, the comparative voltage generating circuit 22 generates one more comparative voltage S12' based upon the inputted reproduction signal S12 input. The comparative voltage generating circuit 22 comprises a low pass filter or an envelope detecting circuit to output the comparative voltage S12'and outputs a threshold voltage for detecting the vicinity of each peak of the reproduction signal S12 as the comparative voltage S12'. Next, the comparator 23 compares the reproduction signal S12 with the comparative voltage S12' and outputs a window signal S16 representing the vicinity of the peak of the reproduction signal S12. An AND gate 26 then gates the peak detecting signal S17 with the window signal S16, thus unnecessary noises included in the peak detecting signal S17 are masked. As the result, the reproduction data signal S18 having only pulses corresponding to "1" in the recorded data signal S11 can be obtained.
However, in the conventional data detecting circuit 3, the level margin of the reproduction signal S12 for the comparative voltage S12' in the comparator 23 is relatively low. Therefore, when a fluctuation in the amplitude of the reproduction signal S12 is caused, the window signal S16 may not be correctly outputted. Furthermore, in the optical memory apparatus such as the above-mentioned magneto-optical disc etc., the reproduction signal S12 includes low frequency components, and therefore, the above level margin is further decreased making the window signal S16 incorrect. Furthermore, for example, when a drop-out is caused in the reproduction signal S12 as indicated by circle F of FIG. 5, the pulse of the window signal S16 is not outputted, and the window is not present. Therefore, the detecting pulse does not appear in the reproduction data signal S18 although a pulse representing the peak appears in the peak detecting signal S17.
As apparent from the above, the conventional data detecting circuit 3 has such a disadvantage such that errors in detecting the reproduction data signal S18 are often caused when a fluctuation is caused in the amplitude of the reproduction signal, low frequency components are included in the reproduction signal, or a drop-out is caused.